On the Influence of Surface Roughness on RADARSAT-2 Polarimetric Observations
Jesus Alvarez-Mozos(1), Maria Gonzalez-Audicana(1) and Javier Casalí(1)
(1) Public University of Navarre, Los Tejos, Arrosadia s/n, 31006, Pamplona, Spain
Surface roughness represents one of the main disturbing factors for the estimation of soil moisture from radar observations. At present, several issues related to surface roughness are not fully understood. On the one hand, due to its strong influence on radar observations, roughness must be known with large detail in order to obtain reasonable moisture estimates. On the other hand, its large spatial variability and multi-scale nature make the acquisition of ground roughness measurements complicated and costly. So far, single-configuration observations have shown a limited ability to perform operative soil moisture estimations. The availability of polarimetric space-borne radar sensors brings new expectations in this field of research. It is believed that polarimetric analysis will lead to a much more complete description of the scattering process and will, hopefully, result in an improved understanding of the roughness interaction and, hence in a better soil moisture estimation ability.
In this study we present the results of an experiment carried out in October 2008, in Navarre, Spain. Six experimental fields were tilled differently to create different roughness classes. Two RADARSAT-2 Fine Quad-Pol observations with differing incidence angles, were acquired over the area with a three days time difference. Coinciding with image acquisitions, detailed soil moisture and roughness ground measurements were acquired using, respectively, a calibrated TDR probe and a 5 m long laser-profilometer.
A complete polarimetric analysis of the images was carried out and the influence of surface roughness on the main polarimetric variables was assessed. The relation between the polarimetric variables observed at different incidence angles was also evaluated in an attempt to separate roughness classes, that could be useful for moisture estimation.